U.S. patent number 5,956,701 [Application Number 08/874,530] was granted by the patent office on 1999-09-21 for method and system for using an artificial neural net for image map processing.
This patent grant is currently assigned to International Business Machines Corporation. Invention is credited to Paul Robert Habermehl.
United States Patent |
5,956,701 |
Habermehl |
September 21, 1999 |
Method and system for using an artificial neural net for image map
processing
Abstract
The present invention provides a method and system for
processing an image, the method comprising the steps of providing
an image and specifying a portion of the image. The present
invention also includes utilizing an artificial neural net to
associate the portion of the image with an application, wherein a
plurality of positions are selected within the specified portion of
the image.
Inventors: |
Habermehl; Paul Robert (Round
Rock, TX) |
Assignee: |
International Business Machines
Corporation (Armonk, NY)
|
Family
ID: |
25364013 |
Appl.
No.: |
08/874,530 |
Filed: |
June 13, 1997 |
Current U.S.
Class: |
706/20 |
Current CPC
Class: |
G06T
7/143 (20170101); G06T 7/11 (20170101); G06T
2207/20092 (20130101); G06T 2207/20081 (20130101); G06T
2207/20084 (20130101) |
Current International
Class: |
G06T
5/00 (20060101); G06F 019/00 (); G06T 001/40 () |
Field of
Search: |
;706/20,27,16 |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
"Pattern Recognition Method for Personal Computers," IBM Technical
Disclosure Bulletin, vol. 38, No. 01, 509-512 (Jan. 1995)..
|
Primary Examiner: Downs; Robert W.
Attorney, Agent or Firm: Sawyer & Associates Salys;
Casimer K.
Claims
What is claimed is:
1. A method for processing an image, the method comprising the
steps of:
a) providing an image;
b) specifying a portion of the image to be associated with an
application; and
c) utilizing an artificial neural net to define the portion of the
image, wherein a plurality of positions are selected within the
portion of the image and wherein the artificial neural net uses the
plurality of positions to define the portion of the image.
2. The method of claim 1, wherein the artificial neural net
receives information related to the plurality of positions to
provide a boundary for the portion of the image.
3. The method of claim 2, wherein the plurality of positions are
selected by clicking a pointing device.
4. The method of claim 2, wherein the plurality of positions are
selected by scribbling a pointing device.
5. The method of claim 1, wherein the application is accessed via
an internet.
6. The method of claim 1, wherein the association is a software
link.
7. A system for using an artificial neural net for image map
processing, the system comprising:
a display for displaying an image, the image including a portion to
be associated with an application, the portion including a
plurality of positions;
a processor coupled to the display;
a server which can be coupled with the processor, the server
including an artificial neural net which can be trained to define
the portion of the image using the plurality of positions.
8. The system of claim 7, wherein the plurality of positions are
selected by clicking a pointing device.
9. The system of claim 7, wherein the plurality of positions are
selected by scribbling a pointing device.
10. The system of claim 7, wherein the application is accessed via
an internet.
11. The system of claim 7, wherein the association is a software
link.
12. A method for processing an image, the method comprising the
steps of:
a) providing an image;
b) specifying a portion of the image to be linked with information
via an internet; and
c) utilizing an artificial neural net to define the portion of the
image, wherein a plurality of positions are selected within the
specified portion of the image and wherein the artificial neural
net uses the plurality of positions to define the portion of the
image.
Description
FIELD OF THE INVENTION
The present invention relates generally to image map processing,
particularly to a system and method for using an artificial neural
net for associating a portion of an image with an application such
as those which can be accessed through an Internet link.
BACKGROUND OF THE INVENTION
As the use of the Internet gains popularity, an unprecedented
number of users are now performing more complex computer functions.
For example, numerous users are creating their own home page for
access via the Internet. As part of their home page, many users are
utilizing internet links from one subject to another subject. A
typical link is, for example, a link which associates a first
application with another application. For example, if the user is
viewing a map of a county, the user may select a city which
includes links to more information regarding the selected city.
When a portion of an image is selected for a link, the portion of
the image which is associated with the linked application must
initially have been properly identified. For example, if a city is
selected to be linked to other information, then the boundaries of
the city typically are required to be defined to allow the user to
select any portion of the city in order to initiate the link.
If the shape of the portion of the image is a simple geometric
shape, such as a city with its boundary in the shape of a
rectangle, then it is relatively simple to identify the boundary
for the portion of the image. However, if the portion of the image
is shaped in a complex shape, then many data points are required to
define the shape of the boundary of the portion of the image. This
identification of the boundary of a selected portion of the image
can be very tedious to the programmer when the programmer attempts
to identify the portion of the image which is to be utilized for
future linking with another application.
Another problem with the conventional method of identifying a
boundary of a portion of an image occurs when a first portion of
the image and a second portion of the image are each respectively
to be linked with a first application and a second application. If
the first and second portions share a boundary, then the
conventional identifying method typically requires the shared
boundary to be defined twice: once for the first portion and
another for the second portion. Accordingly, duplicity of
identification of a boundary is required for this circumstance.
Accordingly, what is needed is a method and system for identifying
a portion of an image to be associated with another application,
wherein the method and system are easier for the programmer to
apply. The present invention addresses such a need.
SUMMARY OF THE INVENTION
The present invention provides a method and system for processing
an image, the method comprising the steps of providing an image and
specifying a portion of the image. The present invention also
includes utilizing an artificial neural net to associate the
portion of the image with an application, wherein a plurality of
positions are selected within the specified portion of the
image.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of a system in which the present
invention can reside.
FIG. 2 is a flow diagram of a conventional method for defining
perimeters of a portion of an image.
FIG. 3 is an example of an image with perimeters of portions of the
image having been defined according to the conventional method
shown in FIG. 2.
FIG. 4 is a flow diagram of a method according to the present
invention for defining a portion of an image.
FIG. 5 is a more detailed flow diagram of a portion of the method
according to the present invention shown in FIG. 4.
FIG. 6 illustrates an example of portions of an image identified
via the method according to the present invention.
DESCRIPTION OF THE INVENTION
The present invention relates to a method and system for image map
processing. The following description is presented to enable one of
ordinary skill in the art to make and use the invention and is
provided in the context of a patent application and its
requirements. Various modifications to the preferred embodiment
will be readily apparent to those skilled in the art and the
generic principles herein may be applied to other embodiments.
Thus, the present invention is not intended to be limited to the
embodiment shown but is to be accorded the widest scope consistent
with the principles and features described herein.
FIG. 1 shows a block diagram of a system in which the present
invention can reside. The system includes a display 102, a keyboard
104, a pointing device 106, a processor 108, a memory 110, a disk
drive 112, and an Internet gateway 116, all coupled together by the
system bus 114.
FIG. 2 shows a flow diagram of a conventional method for defining a
portion of an image for purposes such as providing a link to
another application, such as an Internet site. The image file is
provided via step 200. The parameters for each region are defined
via step 202. Associations, such as Internet links, are associated
to each defined region via step 204. The file of associated links
is then mapped via step 206 which can later be accessed from the
web page via step 208.
FIG. 3 shows an illustration of an example of an image with
portions of the image defined according to the conventional method
shown in FIG. 2. Regions 300a-300e are shown within the image.
Perimeters of the region, such as regions 300a and 300b, are
typically defined by identifying the perimeters with (x,y) inputs
302. If the portion of the image is a simple geometric figure, such
as a rectangle, few (x,y) inputs are required since a straight line
may be drawn between two separate inputs. However, for more complex
images and regions, numerous separate inputs are typically required
to define a perimeter which consists of curves or jagged edges.
Inputting a large number of (x,y) inputs is tedious for the
programmer since the number of (x,y) inputs becomes more time
consuming for a programmer as the number of inputs increases.
Additionally, boundaries 304 which are shared between two regions
which are to be defined typically need to be defined twice: once
for the first region 300a and another time for the second region
300b.
What is needed is a system and method for defining portions of an
image which require fewer inputs and less redundancy. The present
invention addresses such a need.
FIG. 4 shows a flow diagram of a method according to the present
invention for defining a portion of an image for purposes such as
associating the portion of the image with an application such as a
link to an Internet file. The image file is provided via step 400.
The image file can be any image file format including GIF file or
JPEG file. Training data can be generated via step 402, and
provided via step 404. The training data can be generated by
selecting various locations within the selected portion of the
image by, for instance, clicking a pointing device, such as a
mouse, within the portion of the image.
An artificial neural net training tool is then utilized via step
406. Artificial neural net (ANN) training tools are well known to
those skilled in the art. The artificial neural net is trained via
step 408 and the results can be utilized via an application such as
a Java applet via step 410. This information can then later be
accessed via an application such as a web page via step 412.
FIG. 5 shows a more detailed flow diagram of an example of steps
402-408 of FIG. 4. The image, map, or photo is displayed via step
500. A cursor is also displayed via step 502. The user then labels
regions of the image to be created via step 504. The labeling of
regions can be as simple as listing identifiers to be associated to
particular regions. A region to be defined is specified via step
506. An example of specifying a region is to identify one of the
labels the user created via step 504.
The user then defines the specified region by selecting points
within the region by performing an act such as randomly clicking an
input device such as a mouse, associated with a cursor, within the
specified region. The more times the user clicks in the region, the
better defined the region will be. An example of the number of
clicks required will be 3 or 4 clicks to define convex or concave
areas of the region. The clicks utilized in the internal portion of
the region can be sparse while the number of clicks closer to the
boundary of the region is preferred to be more concentrated. The
concentration of the number of clicks in any given region or in a
portion of the region depends on the complexity of the region
itself. For example, a boundary which is more complex will require
a larger number of clicks than a boundary which is relatively
simple. For instance, a portion of a region where the boundary
includes a tight curve next to a series of waves and peaks is
preferred to have a higher density of clicks than a portion of the
region with a substantially straight line. The higher the density
of clicks, the more accurate the definition of the region will
be.
The region can also be identified with training data other than
clicks, such as scribbles caused by the mouse. If a continuous
motion, such as scribbles, are used, then it is preferred to use a
sampling of (x,y) inputs. For instance, one (x,y) input can be used
out of every 10 or 100 (x,y) inputs within a portion of a
scribble.
The user indicates a name of a region to be specified. The user can
then input data, by a method such as clicking a pointing device,
within a region to identify the region which has been named. The
inputs train an artificial neural net which can be located in a
server. The various inputs within a region can be utilized by an
artificial neural net to define the region and store the definition
of the region to later be accessed. An application of the data
derived via the artificial neural net is allowing a user to select
any portion of a region within a map to link to additional
information linked to that region.
After the training data is generated for a specific region via step
508, it is determined if another region is to be defined via step
510. If so, then the region to be defined is specified via step 506
and more training data is generated to define the specified region
via step 508. After all training data is generated, the neural net
algorithm uses the data to train the artificial neural net.
FIG. 6 is an illustration of an example of an image with portions
defined according to the present invention. The image is shown with
various regions 600a-600f with region 600a and 600b having been
defined via inputs 604, which may be derived from clicks of the
mouse connected to the cursor or scribbles within the selected
region of the image. Note that the boundary 602 shared between
defined regions 600a and 600b need not be defined twice since the
data inputs do not need to be exactly along the boundary.
This random input of data within a particular region allows a more
user friendly method of defining a particular region of an image
since the user does not need to identify the exact boundary.
A method and system for using an artificial neural net for image
map processing has been disclosed. Software written according to
the present invention is to be stored in some form of computer
readable medium, such as memory or CD ROM, or transmitted over a
network, and executed by a processor.
Although the present invention has been described in accordance
with the embodiments shown, one of ordinary skill in the art will
readily recognize that there could be variations to the embodiments
and those variations would be within the spirit and scope of the
present invention. Accordingly, many modifications may be made by
one of ordinary skill in the art without departing from the spirit
and scope of the appended claims.
* * * * *